There are a lot of people on the Internet that make up results.
Monero Mining Performance
- Thread starter Patrick
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
does anybody has a i7 8600k that they have tested?
I think it will do around the 483. becauseit's the same CPU as the i7 7700k, but then with 2 extra cores.
i7 7700k is 322(data from the list) /4*6=483. And it has the same l3 cache per core (2mb).
if this is true it is maybe beter then the ryzen 1700. we just need to know how much power it uses and if it's beter to OC (atleast the cache speed then)
I think it will do around the 483. becauseit's the same CPU as the i7 7700k, but then with 2 extra cores.
i7 7700k is 322(data from the list) /4*6=483. And it has the same l3 cache per core (2mb).
if this is true it is maybe beter then the ryzen 1700. we just need to know how much power it uses and if it's beter to OC (atleast the cache speed then)
It may be good. But it's 6 cores + a dimm + mobo + psu + 24x7 cooler + a case and by the time you've got enough of them to be useful you'll have too many nodes to manage and need more switch ports.
AMD EPYC 7401P is a monstrosity of a chip for mining at $1100. A $1400 server for the EPYC is like getting 4 mining nodes so <400 for mobo psu case and cooler and you've got fewer nodes to manage.
yea that true tough, but I don't have the money to buy that. Here in the netherlands you will need to pay 1250 € and around the 600$ for just the motherboard.
and the ROI is around the 1100 days and yes the i7 8700k is a bit more, but I also plannend to use the cpu for gaming and other stuff where single core performance helps more then multi core performance
Hello to Everyone,
and a special hello for Patrick,
I've just done the registration to this forum to ask you a question
are you sure about the perfomance of
1x AMD EPYC 7351P = 1438H/s (32 threads 179W)
? this have double the L3 ram of the threadripper 1950x and almost same perfomance 0_O
all epyc servers with 64mb L3 seems to have quite low performance (almost half of expected ) do you have any detailed information about these tests?
did you tested those servers personally ?
Thank you! and have a nice day
and a special hello for Patrick,
I've just done the registration to this forum to ask you a question
are you sure about the perfomance of
1x AMD EPYC 7351P = 1438H/s (32 threads 179W)
? this have double the L3 ram of the threadripper 1950x and almost same perfomance 0_O
all epyc servers with 64mb L3 seems to have quite low performance (almost half of expected ) do you have any detailed information about these tests?
did you tested those servers personally ?
Thank you! and have a nice day
Hi,
please keep in mind, that Ryzen Threadripper has got a much higher frequency than Epyc 7351P; 3.4 GHz with 4.2 GHz XFR-Turbo vs. 2.4 GHz with 2.9 GHz Turbo.
Besides: due to its 32 MB L3-Cache and the character of Monero (2 MB per thread) the miner works with 16 threads on Threadripper, which is exactly its number of real cores, which is perfect. On Epyc 7351P instead, although it has just 16 cores, as well, the miner works with 32 threads. But logical SMT-cores don't scale as good as real cores. Nevertheless Epyc 7351P is faster than Threadripper, although it has just the same amount of CPU cores and works with lower frequency.
Last edited:
Hi Nero24,
I understand your point but not completely I've already supposed that the core number and core frequency could be a reason for this discrepancy,
anyway also processor have so low hashrate...
1x AMD EPYC 7601 = 1786H/s (196W)
1950x supposing running all cores at 4ghz constant , total freq = 64000Hz x 32mbL3 = 1204H
7351P total physical frequency = 46400Hz x 64mbl3 = 1438H
7601 = 86400Hz x 64mbl3 = 1786H
if the cpu freq is the limiting factor of the 7351p , the 7601 which have way more cpu frequency on 32 physical cores should give a huge advantage over the 7351P but it's not.. , I mean the 7351p having lower cpu freq does a decent job but not very muhc.. ( this let me suppose that the cpu was the bottleneck but then the 7601 double the cpu freq and the step over the 7351P is small.. ) probably the two resources are both close to the limit and both freq and cpu are the limit ? maybe a 7601 with total 128000Hz can reach 2400H/s , ( supposing that the 1950x run always at 4ghz constant )
thank you
- reference :
Epyc - Wikipedia
Ryzen - Wikipedia
I understand your point but not completely I've already supposed that the core number and core frequency could be a reason for this discrepancy,
anyway also processor have so low hashrate...
1x AMD EPYC 7601 = 1786H/s (196W)
1950x supposing running all cores at 4ghz constant , total freq = 64000Hz x 32mbL3 = 1204H
7351P total physical frequency = 46400Hz x 64mbl3 = 1438H
7601 = 86400Hz x 64mbl3 = 1786H
if the cpu freq is the limiting factor of the 7351p , the 7601 which have way more cpu frequency on 32 physical cores should give a huge advantage over the 7351P but it's not.. , I mean the 7351p having lower cpu freq does a decent job but not very muhc.. ( this let me suppose that the cpu was the bottleneck but then the 7601 double the cpu freq and the step over the 7351P is small.. ) probably the two resources are both close to the limit and both freq and cpu are the limit ? maybe a 7601 with total 128000Hz can reach 2400H/s , ( supposing that the 1950x run always at 4ghz constant )
thank you
- reference :
Epyc - Wikipedia
Ryzen - Wikipedia
Well it you can't give a blanket statement like that, i'm running three Wiwynn Open Compute servers with E5-2660's and 2665's and its working for me.
Now granted I bought the servers cheap and the cpu's used and run the least ram I can get away with and hard drives I already had but that was my goal since there was no way I was going to buy the newest servers or the latest cpu's.
Of course if I win the lottery.......
Hi @ffvv using our newer methodology the EPYC parts are doing much better. The constraint in our original runs is that the miner is running across NUMA nodes.
xmrig Monero Dwarfpool Image Testing
xmrig Monero Dwarfpool Image Testing
Entire systems. You are correct in that you are looking for the more informative metric
Hi -
New here. This is a good thread.
I am interested in the Open Compute servers for XMR mining. Trying to decide what CPUs are the best bang for the buck.
In post #1 I see:
2x Intel Xeon E5-2670 V1 = 829H/s
2x Intel Xeon E5-2650 V1 = 750H/s
However, I do not see ANY E5-26xx v2 CPUs. Several v1, v3, v4, but no v2. I am wondering why. Is that a really bad CPU for mining?
For example, I am looking at the E5-2660 v2 - would that not be quite a bit better than the v1? It has 25MB L3 cache so should be able to run 12 threads. Anyone care to guess the H/s given the v1 numbers above?
Appreciate any advice on the best Open Compute server CPU to start out with (probably in one of the Quanta v2 systems).
Thanks!
New here. This is a good thread.
I am interested in the Open Compute servers for XMR mining. Trying to decide what CPUs are the best bang for the buck.
In post #1 I see:
2x Intel Xeon E5-2670 V1 = 829H/s
2x Intel Xeon E5-2650 V1 = 750H/s
However, I do not see ANY E5-26xx v2 CPUs. Several v1, v3, v4, but no v2. I am wondering why. Is that a really bad CPU for mining?
For example, I am looking at the E5-2660 v2 - would that not be quite a bit better than the v1? It has 25MB L3 cache so should be able to run 12 threads. Anyone care to guess the H/s given the v1 numbers above?
Appreciate any advice on the best Open Compute server CPU to start out with (probably in one of the Quanta v2 systems).
Thanks!
Thank you for the very quick replies guys!
Any other CPUs I should be looking at instead of the E5-2660 v2 for a low cost setup? I would like to start at a relative low cost so that rules out EPYC and later gen Intels I think and makes me look at the Open Computer servers which as far as I know only supports up to E5-26xx V2 (max. 115W TDP?).
Any other CPUs I should be looking at instead of the E5-2660 v2 for a low cost setup? I would like to start at a relative low cost so that rules out EPYC and later gen Intels I think and makes me look at the Open Computer servers which as far as I know only supports up to E5-26xx V2 (max. 115W TDP?).
@kfriis
Take a look at this thread , it may answer some of your questions.
Dedicated Crypto Mining Open Compute server build.
https://forums.servethehome.com/ind...rypto-mining-open-compute-server-build.15021/
Do you have a budget in mind?
How much do you pay for power?
Take a look at this thread , it may answer some of your questions.
Dedicated Crypto Mining Open Compute server build.
https://forums.servethehome.com/ind...rypto-mining-open-compute-server-build.15021/
Do you have a budget in mind?
How much do you pay for power?
Thanks @Marsh
I have been reading a lot of threads here lately but I am now trying to narrow everything down so I can move forward - hence the questions!
I think, as a rough guideline, about USD1000 would be my initial budget. If I can find a "good deal", this number can be somewhat flexible. I pay about 10c/kWh for power, but might have options to bring that down in the future - that number is what I currently pay.
Appreciate any help -- from reading here for some time, I have been impressed by the knowledge of the frequent posters like yourself.
I have been reading a lot of threads here lately but I am now trying to narrow everything down so I can move forward - hence the questions!
I think, as a rough guideline, about USD1000 would be my initial budget. If I can find a "good deal", this number can be somewhat flexible. I pay about 10c/kWh for power, but might have options to bring that down in the future - that number is what I currently pay.
Appreciate any help -- from reading here for some time, I have been impressed by the knowledge of the frequent posters like yourself.